Results The genotype distribution of the polymorphic CYP2A6 and GSTP1 in controls and cases is shown in table 1. Both the genotype and allele frequencies of GSTP1 in controls were found to be in Hardy-Weinberg equilibrium. As evident from table 1, the distribution of CYP2A6 genotypes in HNSCC cases was significantly different from that in controls, indicating a relationship between CYP2A6 genotypes and HNSCC risk. The frequency of genotypes with one or both deletion alleles (*4C) of CYP2A6 (*1A/*4C, *1B/*4C, *4C/*4C) was significantly higher in controls when compared to the cases resulting in a marked decrease in HNSCC risk particularly in cases with *4C/*4C genotype (O.R. 0.16; 95% C.I. 0.03-0.78) which was statistically significant (p=0.01). This decrease in risk (O.R. 0.20; 95% C.I. 0.05 – 0.82) persisted even when the data was adjusted for age, smoking, daily consumption of alcohol and tobacco chewing in multivariate logistic regression and was also found to be statistically significant. However, no association of *1A/*1B and *1B/*1B was observed with HNSCC risk. As observed with CYP2A6, the increase in the frequency of variant genotypes of GSTP1 in the controls when compared to the cases resulted in a decrease in the odds ratio and the decrease in risk was also significant after adjusting for age, smoking, alcohol and tobacco consumption. Likewise, combination of variant genotypes of CYP2A6 and GSTP1 was over-represented amongst the controls (22%) when compared to the cases (16%) resulting in significant decrease in HNSCC risk (O.R. 0.48; 95% C.I. 0.3 – 0.77) (Table 2).
The effects of the interaction of the risk modifiers i.e. cigarette smoking or tobacco-chewing with the variant genotypes of CYP2A6 is summarized in table 3. The frequencies of the variant genotypes of CYP2A6 were increased in the controls (62.4%) when compared to the cases (51%) with the history of smoking which resulted in a significant decrease in HNSCC risk in smokers (O.R. 0.62; 95% C.I. 0.38 – 0.99). Similarly, amongst tobacco-chewers, overrepresentation of variant genotypes in controls (63%) when compared to the cases (51%) also resulted in a significant decrease in HNSCC risk (O.R. 0.61; 95% C.I. 0.38 – 1.0). However, no association with HNSCC risk was observed with variant genotypes in non-smokers and non-tobacco chewers.